DEVELOPMENT AND VALIDATION OF A COARSE-GRAINED MODEL FOR THE ...

45 Pages·2016·1.08 MB·English

DEVELOPMENT AND VALIDATION OF A COARSE-GRAINED MODEL FOR THE
BIOCOMPATIBLE POLYMER POLYCAPROLACTONE
By
ABHINAV SANKARA RAMAN
A thesis submitted to the
Graduate School – New Brunswick
Rutgers, The State University of New Jersey
In partial fulfillment of the requirement
For the degree of
Master of Science
Graduate Program in Chemical and Biochemical Engineering
Written under the direction of
Dr. Yee C. Chiew
and approved by
_____________________________
_____________________________
_____________________________
_____________________________

New Brunswick, New Jersey
May, 2016
ABSTRACT OF THE THESIS

DEVELOPMENT AND VALIDATION OF A COARSE-GRAINED MODEL FOR THE
BIOCOMPATIBLE POLYMER POLYCAPROLACTONE
by ABHINAV SANKARA RAMAN
Thesis Director:
Dr. Yee C. Chiew

Polycaprolactone (PCL), is a biocompatible polyester with many applications towards the
betterment of human health. The advent of superior computational power and novel
simulation techniques such as coarse-grained molecular dynamics simulations, enable the
in-silico study of processes involving nano-polymeric materials over large spatio-temporal
scales. In this study, a coarse-grained model for PCL was developed within the framework
of the MARTINI coarse-grained (CG) force field. The non-bonded interactions were based
on the existing MARTINI bead types, while the bonded interactions were mapped from the
OPLS-UA/AA rendition of PCL. The developed model accurately reproduced the
structural and dynamic properties of the PCL homopolymer, also showing reasonably good
temperature and solvent transferability. We also studied the self-assembly of MePEG-b-
PCL linear diblock copolymers using an existing MARTINI model for MePEG (Methoxy
Polyethylene glycol), by analyzing the critical micelle concentration, as well as the shape,
size and morphology of the nano-polymeric micelles. We obtained excellent agreement of
the critical micelle concentration, while the size was under-predicted compared to
experimental data.
ii
Acknowledgements

First and foremost, I would like to thank my advisor Dr. Yee Chiew, for taking me under
his tutelage. It was both a privilege and honor to work under someone I have always
considered as one of my heroes and inspiration, with respect to my scientific pursuits.
Science aside, I would also like to thank him for imbibing in me his impeccable research
ethics and character, which I’m certain will stay with me for the rest of my life. I would
also like to thank Dr. Aleksey Vishnyakov for his valuable input and discussions with
respect to this project.

I would like to thank my family for their support over all these years. This would not have
been possible without their help. I would like to thank all my friends, especially Subhodh
and Srinivas, for all their help over these three years. I would like to thank J.R.R Tolkien
for writing the Lord of the Rings, whose heroes have inspired and lifted my spirit on several
occasions. Finally, I would like to thank God, for at the end of the day, none of this would
have happened if not for his unrequited love.

iii
Table of Contents

ABSTRACT OF THE THESIS ..................................................................................... ii
Acknowledgements ...................................................................................................... iii
List of Tables ................................................................................................................ vi
List of Figures ............................................................................................................ viii
Chapter 1 Introduction ................................................................................................... 1
Chapter 2 Forcefield Development ................................................................................ 4
Chapter 3 Computational Details ................................................................................... 8
3.1 Atomistic Simulations .......................................................................................... 9
3.2 Coarse-grained Simulations ................................................................................. 9
3.3 Implicit Solvent Simulations .............................................................................. 10
Chapter 4 Results and Discussion ................................................................................ 11
4.1 PCL homopolymer: Structural and conformational properties .......................... 11
4.2 PCL homopolymer: Temperature transferability of the model .......................... 13
4.3 PCL homopolymer: Dynamics of PCL in solution ............................................ 15
4.4 PCL homopolymer: Solvent transferability of the model .................................. 16
4.5 Copolymer model: Self-assembly of amphiphilic diblock copolymers ............. 19
4.5.1 Morphology, size and shape of the self-assembled micelles .................... 20
4.5.2 Critical micelle concentration (CMC) and hydrodynamic radius ............. 25
Chapter 5 Implicit solvent simulations: Transferability and Validation ...................... 28
5.1 Introduction ........................................................................................................ 28
5.2 Transferability to the “Dry” MARTINI forcefield ............................................. 29

iv
5.3 Conformation of PCL homopolymers: “Dry” vs “Wet” MARTINI .................. 30
Chapter 6 Summary and Conclusion ........................................................................... 32
References .................................................................................................................... 34

v
List of Tables
Table 1. Parameters of the bonded interactions in the CG model ................................. 7
Table 2. Time averaged end to end distance of PCL homopolymers in water
at 300K ......................................................................................................................... 11
Table 3. Time averaged radius of gyration of PCL homopolymers in water
at 300K ......................................................................................................................... 12
Table 4. Time averaged principal moments of the gyration tensor of PCL
homopolymers in water at 300K .................................................................................. 13
Table 5. Time averaged radius of gyration of PCL homopolymers in water over a range
of temperatures............................................................................................................. 14
Table 6. Hydrodynamic radius of PCL homopolymers in water ................................. 16
Table 7. Time averaged end to end distance of PCL homopolymers in
acetone-water mixture .................................................................................................. 18
Table 8. Time averaged radius of gyration of PCL homopolymers in
acetone-water mixture .................................................................................................. 19
Table 9. Mean cluster size, radius of gyration and asphericity of the clusters
(MePEG -b-PCL ) at 300K ........................................................................................ 24
17 3
Table 10. Mean cluster size, radius of gyration and asphericity of the clusters
(MePEG -b-PCL ) at 310K ........................................................................................ 24
17 3
Table 11. Mean cluster size, radius of gyration and asphericity of the clusters
(MePEG -b-PCL ) at 310K ........................................................................................ 24
17 2
Table 12. Hydrodynamic diameter of the largest clusters at 310K.............................. 25
vi
Table 13. Critical micelle concentration for the MePEG -b-PCL system
17 2
at 310K ......................................................................................................................... 27
Table 14. Time averaged radius of gyration of PCL homopolymers at 300K
(“Dry” vs “Wet” MARTINI) ....................................................................................... 31
Table 15. Time averaged principle moments of the gyration tensor of PCL
homopolymers at 300K (“Dry” vs “Wet” MARTINI) ................................................ 31

vii
List of Figures
Figure 1. Mapping scheme and the corresponding MARTINI-CG rendition of
PCL, MePEG and MePEG-b-PCL ................................................................................. 5
Figure 2. PCL bond and angle distribution, MePEG-PCL interconnecting beads
bond and angle distribution ............................................................................................ 6
Figure 3. Radius of gyration versus number of monomeric repeat units for PCL
homopolymer in water ................................................................................................. 13
Figure 4. Snapshots of the self-assembly of the MePEG -b-PCL system at 300K,
17 3
through the course of the simulation ............................................................................ 21
Figure 5. Time evolution of the number of clusters for the MePEG -b-PCL system
17 3
at 300K ......................................................................................................................... 22
Figure 6. Morphology of the biggest cluster at 300K .................................................. 23
Figure 7. Time evolution of the number of free monomers and aggregation number of the
clusters for the MePEG -b-PCL system at 310K ...................................................... 26
17 2
Figure 8. PCL bond and angle distributions between “Dry” and “Wet” MARTINI ... 29

viii
1

Chapter 1
Introduction
Poly--caprolactone (PCL), an FDA approved biodegradable hydrophobic polyester, has
been used widely in the biomaterials arena, showing a resurgence in the last decade. PCL,
has known applications in the medical devices field ranging from sutures and wound
dressings to dentistry. However, it is the birth and boom of tissue engineering that has re-
vitalized the importance of PCL as a biomaterial. Due to its far superior rheological
properties, cost-efficient production and longer degradation time, it is now widely applied
as long term implants and scaffolds with huge potentials for bettering human health. In
addition to these, PCL has always been used in nanoparticle based drug delivery systems,
making it one of the few biomaterials with versatile applications1.

Its use in nanoparticle based drug delivery devices is one of potential importance in
treatment of diseases and ailments such as cancer, owing to novel properties such as
targeted delivery and controlled release of drugs2, 3. In a large number of cases, PCL is
functionalized with ligands or other functional groups, and the resulting copolymers are
used for drug delivery. Specifically, amphiphilic copolymers formed by functionalizing
PCL with a hydrophilic functional group, and the block copolymer micelles that they form,
have shown to evade the reticuloendothelial system, thereby providing a much improved
circulation time for the nano-carriers4, 5. Of the many amphiphilic systems studied with
PCL, the polyethylene glycol block polycaprolactone (PEG-b-PCL) is one that has been
widely applied6, 7.
2

Over the last decade, the advent of superior computational power, and the use of mesoscale
simulation techniques such as coarse-grained molecular dynamics simulations, have aided
in a large number of computational studies on the behavior of polymeric molecules,
especially amphiphiles, such as diblock star copolymers8, peptides9, and triblock
copolymers10 to name a few. A recent review paper by Thota and Jiang11, on computational
studies involving amphiphilic materials used specifically for drug delivery applications,
goes to show the importance of developing robust and accurate models for such systems.
Specifically with respect to PCL and the diblock PEG-PCL copolymer, it is worth noting
the CG model of Loverde et al.12, which reproduces the phase diagram of the PEG-PCL
system. Further, the authors also studied the effect of the shape of the nanoparticle on drug
delivery, by analyzing the loading of the popular anti-cancer drug Taxol into self-
assembled nanoparticles. While there are several other standalone CG models for a large
number of polymers, including other amphiphiles, transferability of CG models is
problematic and requires a systematic development across different classes of compounds.

In this study, we have developed a model for PCL within the framework of the popular
MARTINI coarse-grained (CG) force field13, 14. The MARTINI force field, originally
developed for lipids, is a systematically parametrized CG biomolecular force field suitable
for molecular dynamics simulations of a large number of biomolecules. While molecular
models for a large number of molecules such as lipids, proteins and sugars have been
developed for use with the MARTINI force field, only a few polymers have been covered.
The prominent examples include polyethylene oxide (PEO) and polyethylene glycol (PEG)
by Lee et al.15, Polystyrene by Rossi et al.16, PAMAM by Lee and Larson17 and more